Liner Regression Help

Dec 2015
4
0
mississippi
I am doing a medical study for my thesis concerning the number of organs a donor (numbers ranging from from 0-7) will produce given certain circumstances. I am monitoring their blood sugar, ejection fraction, etc (which are whole numbers) but I also have categorical figures such as Race, Gender, etc. On top of that, I have some yes or no questions as well. I can convert these to binary numbers, but would that mess up a multiple regression or give me bad results. Should I look into another model?
 

romsek

MHF Helper
Nov 2013
6,836
3,079
California
What are you trying to show?

This doesn't sound like a regression problem but a grouping problem and for that you use ANOVA.
 
Dec 2015
4
0
mississippi
So basically I am trying to see if a high dosage of steroids or a low dosage of steroids during a case leads to a higher yield of recovered organs. I have other factors too and I can see if any of those are related as well. Mainly what I want is a low dosage=more organs
 

romsek

MHF Helper
Nov 2013
6,836
3,079
California
So basically I am trying to see if a high dosage of steroids or a low dosage of steroids during a case leads to a higher yield of recovered organs. I have other factors too and I can see if any of those are related as well. Mainly what I want is a low dosage=more organs
well I'd start by isolating those two pieces of info, i.e. steroid dosage level and yield of recovered organs, and doing a regression on those and seeing how good your fit is.

If you have a very good fit there's no need to start bringing in other pieces of info. If you don't then what you can do is subdivide your data based on other info, age, race, sex, whatever, and perform individual regressions on those data subsets.

That's how I would start.
 
Dec 2015
4
0
mississippi
You know I totally went for the kill without doing things individually first. Thanks for the help